Hi Samuel!
Simply put, a "cat /dev/zero > [home on nfs]/tmpfile" on a SRS almost
stalls the server. (ls takes >1s!)
It happens on any computer that has its NFS ressource unavailable or
saturated.
Is this really the case? Disk, Memory, CPU, Network, all can be shared
by many users fairly, but NFS can't?
Our users tend to work with very large files, so reading or writing a
couple of GB at a time does happen frequently enough to be a real problem.
"Reading or writing a couple of GIGAbits" at a time has to take some
time, you know.
Of course it takes a while to write say a 4GB file via NFS. That's
exactly my problem. During those 40-50 seconds, all other users
experience severe lag.
but it would definitely help if you had a SAS disks server (faster than
SATA!), using 15k disks
We do have those. The NFS server has no difficulties reading or writing
>500MB/s. The bottleneck is the network interface on the client.
Trunking won't help, obviously, and 10GbE switches are still painfully
expensive.
Also, it's no workaround but you can have a look at NFS mounting
options, like rsize, wsize, syncronous
i/o and so on...
None of those options will make the NFS client layer provide what I need
and it appears it can't provide: Fair sharing between users on the same
host. :/
Elmar
_______________________________________________
SunRay-Users mailing list
[email protected]
http://www.filibeto.org/mailman/listinfo/sunray-users
